In this paper, a direct power control (DPC) approach is used to control the active and reactive power injection into the grid from a photovoltaic (PV) unit. The dynamics of active and reactive power are used to develop the dynamical model of the three-phase grid-connected PV (GCPV) system while a nonlinear adaptive technique is employed to design the proposed controller. The parametric uncertainties are incorporated within the dynamical model of three-phase GCPV systems by considering the parameters as totally unknown. The unknown parameters are estimated through adaptation laws and control Lyapunov functions (CLFs) are formulated during the different stages of the controller design process to ensure the overall stability of the GCPV system. The theoretical stability of the three-phase GCPV, with the proposed control scheme, is analyzed through the negative semi-definiteness of the derivatives of CLFs. Simulations are carried out to further validate the performance of the proposed control scheme on a three-phase GCPV system in terms of injecting power (both active and reactive) into the grid along with the power quality. Simulation results demonstrate effectiveness of the proposed controller as compared to an existing sliding mode controller (SMC).